Spatial Data Analysis

Evgeny Noi

August 26, 2021

Possibly a first use case of spatial analysis

Dr John Snow and Cholera Outbreak

One of the first examples of spatial analysis is from epidemiology. Dr John Snow investigated the cholera outbreak in 1854 in London. He postulated that cholera was transmitted through drinking water and not air. He noticed that the outbreak was centered on Broad Street public drinking pump.

Repeating the first spatial analysis with current tools

There are many more projects…

Stages in Spatial Data Analysis

1. Map the variable you are researching

  • choropleth map
  • flow map
  • cartogram
  • dot density map
  • isoline map
  • …

  • Which areas (counties) have the highest mortality?
    • Why is Chicago significant and Minneapolis is not?
    • What happens in Montana?
    • What happens along the coast?

Analyzing attribute information

Remember normal distributions?

Histograms

Scatter plot

Is there a correlation? (fit the line)

Examples of Spatial Operations

North Carolina Data

## Linking to GEOS 3.8.0, GDAL 3.0.4, PROJ 6.3.1

Reproject

Calculate distance

Distance Matrix

## Units: [m]
##           [,1]     [,2]      [,3]     [,4]
## [1,]      0.00 312184.9 128341.85 475623.3
## [2,] 440561.15 114939.7 590434.80      0.0
## [3,]  18944.03 352719.1  78756.89 517527.8

Overlay

Overlay Analysis

Overlaying in NC

##       [,1]  [,2]  [,3]  [,4]
## [1,]  TRUE  TRUE FALSE FALSE
## [2,]  TRUE  TRUE FALSE FALSE
## [3,] FALSE FALSE FALSE FALSE
## [4,] FALSE FALSE FALSE FALSE

Choropleth Map of County Areas

Choropleth Map of Births in North Carolina

More beautiful and custom maps

Centroids

Centroids + Proportional Symbols

Centroids + Proportional Symbols (Size) + Hue

Major US cities

Count cities in each county

  • Cities are points, counties are polygons
  • Find points in each polygon
  • Tally points
  • Use graduated symbols map

Cities in NC

Cities in NC (choropleth)

Here is how you would do it in ArcGIS Pro:

  • Go to Geoprocessing and use Summarize Within tool. Find instructions here.

Example 2: Look at hazardous waster disposal near cities in NC

Size of NC cities

Hazardous waste

## Reading layer `Hazardous_Waste_Sites' from data source 
##   `https://raw.githubusercontent.com/barguzin/spatial_analysis/master/haz.geojson' 
##   using driver `GeoJSON'
## Simple feature collection with 2577 features and 22 fields
## Geometry type: POINT
## Dimension:     XY
## Bounding box:  xmin: -84.02775 ymin: 33.89647 xmax: -75.60355 ymax: 36.53089
## Geodetic CRS:  WGS 84
##     LOC_COUNTY   n
## 1  MECKLENBURG 321
## 2     GUILFORD 243
## 3         WAKE 240
## 4      FORSYTH 139
## 5       DURHAM 121
## 6  NEW HANOVER  78
## 7     BUNCOMBE  77
## 8      CATAWBA  69
## 9   CUMBERLAND  60
## 10      GASTON  60

Analysis outline

  • Select only NC cities>50,000 population
  • We cannot use degree projection, so we will need UTM
  • Create a 15km buffer around these cities
  • when we create a buffer from point layer (red circles) we get a polygon layer (black circles)

Combine buffers into one shape

Plot hazard sites on a map

Count the number of haz waste disposal sites by these buffer zones

Plan for next week

Plan

  • Submit your Project Data (project panel) by Monday midnight
  • Browse the ArcGIS spatial analysis projects and find something similar to your project